A Partial Discharge Recognition Method and System Architecture for IoT Ring Main Unit

In view of the current situation that the operation environment of the ring main unit of the Internet of Things is harsh and the partial discharge phenomenon is serious, this paper proposes a partial discharge identification method for the ring main unit of the Internet of Things, and designs a partial discharge identification system architecture based on cloud-edge collaboration. This paper first studies the collection and processing of partial discharge signals, and then builds a local Transformer model in the cloud to judge the partial discharge. The partial discharge signals are input into the model, and after the depth feature extraction and transformation are completed through the multilayer self-attention mechanism and feed-forward neural network, the discrimination of the partial discharge mode of the IOT ring main unit is realized. Finally, the experimental results show that the proposed partial discharge identification method and system have high identification efficiency.